Decomposition and Parallel Computation of Inter-provincial Medium and Long-term Transaction Optimization Model Considering Time Coupling Constraints

With the accelerated development of China's unified electricity market construction, inter-provincial medium and long-term transactions are gradually shifting towards higher frequencies, and the minimum time periods for transactions will be refined from the current peak, flat, and valley periods to an hourly basis. In addition, temporal coupling constraints, specifically transmission power ramping constraints, have been introduced to further reflect the physical characteristics of power grid operation in the market trading process. However, the expansion of market membership and the increase in the number of time periods have significantly increased the scale of optimization variables and constraints, leading to substantial challenges in solving the problem. This paper first proposes an inter-provincial transaction optimization model based on transaction path matching, which considers the temporal coupling constraints. Furthermore, the diagonal quadratic approximation method is employed to relax the temporal coupling constraints, transforming the original problem into a parallelizable structure. By decomposing the long-term transaction optimization problem into smaller-scale subproblems for each time period, the model solution is accelerated through parallel computation, resulting in a significant improvement in computational efficiency. Numerical results demonstrate the effectiveness of the proposed approach.